The effects of ambient temperature, humidity and season of year on urine composition in patients with nephrolithiasis

Authors


Brian H. Eisner, Department of Urology, GRB 1102, 55 Fruit Street, Boston, MA 02114, USA. e-mail: beisner@partners.org

Abstract

Study Type – Prognosis (cohort series)

Level of Evidence 2b

What's known on the subject? and What does the study add?

Epidemiologic studies have shown that warmer climates are associated with increased incidence of nephrolithiasis. Many hypothesize that this is due to dehydration and lower urine volumes. The current study of stone formers reports that greater temperatures are associated with significant increases in urine calcium which may shed light on the mechanism underlying the increased stone incidence associated with increased ambient temperature.

OBJECTIVE

  • • To understand the effects of temperature, humidity and season of year on 24-h urine composition in patients with nephrolithiasis.

PATIENTS AND METHOD

  • • A retrospective review was performed of patients evaluated at four metabolic stone clinics.
  • • Multivariate linear regression models examined the relationship between mean temperature, average humidity, season of year and 24-h urine composition.
  • • Multivariate models adjusted for known risk factors for stone disease.
  • • Mean temperature and average humidity data were obtained from http://www.weatherunderground.com based on patient-provided addresses.

RESULTS

  • • A total of 599 patients were included in the study, comprising 239 women and 360 men with a mean age of 53.6 years (sd 15.0).
  • • Mean temperature was 16.9 °C (sd 4.8, range −21.1 to 38.3 °C) and average humidity was 58.1% (sd 23.5, range 11–100%).
  • • On multivariate linear regression, increasing temperature was associated with increasing urine calcium (β= 11.3, 95% CI 2.2–20.0), super-saturation of calcium oxalate (β= 0.6, 95% CI 0.2–0.9), super-saturation of calcium phosphate (β= 0.14, 95% CI 0.03–0.2), and decreasing urine sodium (β=−5.2, 95% CI −10.3 to −0.1).
  • • As seasons become warmer (i.e. from winter to autumn to spring to summer), changes were increased urine volume (β= 0.09, 95% CI 0.01–0.2) and decreased super-saturation of calcium phosphate (β=−0.2, 95% CI −0.3 to −0.03).
  • • There were no associations between quintile of humidity and any 24-h urine constituents.

CONCLUSIONS

  • • Increasing temperature may increase stone risk by increasing urine excretion of calcium, and the super-saturation of calcium oxalate and calcium phosphate.
  • • These findings were independent of humidity and of season of year.
  • • This appears to be related to a physiological impact of temperature itself, rather than to geographic location.

INTRODUCTION

Nephrolithiasis is a common cause of morbidity in the USA, with a lifetime prevalence of 5–10% [1,2]. Several temperature-related risk factors are associated with the development nephrolithiasis – among them are hot climates [2,3], occupations which may expose individuals to heat exposure or dehydration [3], and poor oral fluid intake [4]. Several studies have examined the relationship between season (i.e. summer vs winter) and either the incidence of nephrolithiasis or 24-h urine composition, with mixed results [5–7]. Most reports suggest that dehydration or decreased 24-h urine volume is the driving force behind kidney stone formation. We hypothesize that the physiological effects of the ambient temperature itself, not urine volume may be related to kidney stone risk in stone-forming patients. The current study examines the relationship between temperature, humidity, season of year and 24-h urine composition in a cohort of patients with nephrolithiasis in a cross-geographical survey.

MATERIALS AND METHODS

Institutional review board approval was obtained before the initiation of this study. A retrospective review was performed of 24-h metabolic urinalysis databases from four metabolic stone clinics (Boston, MA; Lebanon, NH; Nashville, TN; Scottsdale, AZ, USA). Patients with known stone disease who underwent metabolic evaluation of recurrent nephrolithiasis and who were ≥18 years of age were identified for study inclusion. Outpatient clinic and hospital records, as well as 24-h urine composition data were analysed. Body mass index was calculated as weight in kilograms divided by the square of height in metres (kg/m2) from self-reported patient height and weight values at the time of 24-h urine collection. Electronic medical records were reviewed to determine patients' demographic information (self-reported race, gender), medical history (i.e. history of diabetes mellitus, hypertension, gout) and medication use (i.e. thiazide diuretics, potassium citrate, allopurinol). Patients were excluded if body mass index or past medical history could not be obtained, or if their 24-h urine collections were deemed to be inadequate (for men, 24-h urine creatinine <800 mg; for women, 24-h urine creatinine <600 mg). Mean temperature and average humidity for the date of the 24-h urine collection were recorded from http://www.weatherunderground.com based on patient-provided addresses.

Subjects who presented to the clinics at all four sites for metabolic stone evaluation underwent 24-h urinalyses by the same commercial laboratory (Litholink®, Chicago, IL, USA). Self-reported height and weight were recorded at the time of urine collection. Standard urinary parameters were evaluated, including sodium, calcium, citrate, creatinine, uric acid, oxalate, potassium, phosphorus, magnesium, sulphate, pH, urea nitrogen and urine volume. Super-saturation ratios of calcium oxalate, calcium phosphate and uric acid were calculated using the iterative computer program equil 2.

In our analysis, we included only a single 24-h urinalysis. For patients with more than one 24-h urinalysis available, only the first urinalysis was used. Multivariate linear regression examined the relationship between the quintile of average temperature, quintile of mean humidity, season and 24-h urine composition. Season was treated as a continuous variable (1 = winter, 2 = autumn, 3 = spring, 4 = summer) for purposes of multivariate linear regression. Regression models adjusted for known risk factors for stone disease (age, gender, body mass index, hypertension, diabetes mellitus, gout, relevant medications and 24-h urine constituents). Studentized residuals were calculated to ensure linear relationships and avoidance of type I errors. All tests were two-sided with significance set at P < 0.05. The 95% confidence intervals were calculated for all regression coefficients. All analyses were performed using JMP-SAS 8.0. (SAS Institute, Cary, NC, USA).

RESULTS

In all, 599 patients from four metabolic stone clinics met the inclusion criteria. Female to male ratio was 239 : 360 (i.e. 39.9% were women), mean age was 53.6 years (sd 14.9), mean body mass index was 28.2 (sd 6.2). Hypertension was present in 202 patients (33.7%), diabetes mellitus was present in 65 patients (10.9%), and gout was present in 24 patients (4.0%). Mean temperature was 16.9 °C (sd 4.8, range −21.1 to 38.3 °C) and average humidity was 58.1% (sd 23.5, range 11–100). Mean temperature and average humidity for each site included in the study are shown in Table 1. Season of year for 24-h urine collection was winter in 15.1% of patients, autumn in 14.2% of patients, spring in 23.8% of patients, and summer in 46.7% of patients.

Table 1. Mean temperature and average humidity for each site
LocationNo. of patients (% of total)Temperature (°C) mean (sd)Humidity (%) average (sd)
Boston, MA305 (51.0)11.6 (8.8)69.2 (16.5)
Lebanon, NH100 (16.7)7.8 (6.6)71.7 (14.3)
Nashville, TN27 (4.5)10.1 (9.3)65.0 (13.4)
Scottsdale, AZ167 (27.8)32.9 (14.6)28.2 (11.4)

On multivariate linear regression, increasing temperature quintile was associated with increasing urine calcium (β= 11.3, 95% CI 2.2–20.0), super-saturation of calcium oxalate (β= 0.6, 95% CI 0.2–0.9), super-saturation of calcium phosphate (β= 0.14, 95% CI 0.03–0.2), and decreasing urine sodium (β=−5.2, 95% CI −10.3 to −0.1). As seasons become warmer (i.e. from winter to autumn to spring to summer), changes were increased urine volume (β= 0.09, 95% CI 0.01–0.2) and decreased super-saturation of calcium phosphate (β=−0.2, 95% CI −0.3 to −0.03). There were no associations between quintile of humidity and any 24-h urine constituents. These results are shown in Table 2. For all significant values, Studentized residuals indicated a linear relationship.

Table 2. Multivariate linear regression examining the relationship between temperature, humidity, season and 24-h urine composition
 Temperature quintile β (95% CI)Humidity quintile β (95% CI)Season β (95% CI)
  • *

    P < 0.05. 95% CI, 95% confidence interval; UUN, urine urea nitrogen; Cr, creatinine; SSCaOx, super-saturation of calcium oxalate; SSCaP, super-saturation of calcium phosphate; SSUa, super-saturation of uric acid. Multivariate models adjusted for age, gender, co-morbidities (diabetes, hypertension), medications (thiazide, potassium citrate), and 24-h urine constituents.

Volume (L/day)−0.02 (−0.09 to 0.04)−0.01 (−0.06 to 0.03)0.09* (0.01 to 0.17)
Calcium (mg/day)11.3* (2.23 to 20.3)5.59 (−0.4 to 11.6)−2.87 (−13.6 to 7.9)
Oxalate (mg/day)1.5 (−0.07 to 3.1)−0.3 (−1.4 to 0.7)0.6 (−1.3 to 2.5)
Citrate (mg/day)17.8 (−11.0 to 46.6)−7.6 (−26.8 to 11.6)−17.0 (−51.1 to 17.2)
pH0.005 (−0.04 to 0.05)0.01 (−0.01 to 0.04)−0.03 (−0.08 to 0.02)
Uric acid (g/day)0.002 (0.01 to 0.02)0.002 (−0.007 to 0.01)0.01 (−0.006 tp 0.03)
Sodium (mmol/day)−5.2* (−10.3 to −0.06)−0.43 (−3.9 to 3.0)2.0 (−4.0 to 8.1)
Magnesium (mg/day)0.17 (−3.3 to 3.7)−1.6 (−3.9 to 0.75)−1.3 (−5.4 to 2.9)
UUN (mg/day)−0.02 (−0.2 to 0.2)0.14 (0.02 to 0.3)−0.03 (−0.2 to 0.2)
Cr (mg/day)1.2 (−21.4 to 23.7)−1.1 (−16.1 to 13.9)−8.8 (−35.5 to 17.8)
SSCaOx0.6* (0.2 to 0.9)0.02 (−0.2 to 0.3)−0.3 (−0.7 to 0.08)
SSCaP0.14* (0.03 to 0.2)0.06 (−0.008 to 0.12)−0.2* (−0.3 to −0.03)
SSUa−0.05 (−0.14 to 0.04)−0.02 (−0.08 to 0.04)0.07 (−0.04 to 0.2)

A sub-analysis of the stone clinics located in northeastern USA (Boston, MA and Lebanon, NH) was performed. This sub-group comprised 312 patients. On multivariate linear regression, increasing temperature was associated with significant increases in urine calcium (β= 0.96, 95% CI 0.47–1.44), oxalate (β= 0.11, 95% CI 0.04–0.18), super-saturation of calcium oxalate (β= 0.03, 95% CI 0.009–0.04). There were no associations with temperature and any other 24-h urine constituents.

DISCUSSION

Nephrolithiasis, with a lifetime prevalence of 5–10%, is a common cause of morbidity and has been linked to climate in previous studies [1,2]. Stamatelou et al. [2] observed a 6.6% prevalence of nephrolithiasis in the southern USA compared with a 3.3% prevalence in the more temperate western USA. In addition, season of year has been linked to incidence and prevalence of nephrolithiasis with the hottest months of the year being notable for increased development of kidney stones [8,9]. Studies of urine composition have yielded varied results: several reports describe an association between season and urine oxalate [5–7], urine calcium [5,6], with urine oxalate and calcium excretion being significantly increased in summer compared with winter. Conversely, other studies have failed to show such associations [10].

We hypothesize that temperature itself is related to 24-h urine composition and performed a retrospective study to examine this relationship. We carefully designed our multivariate analysis to include co-variates such as temperature, humidity and season of year to determine the relationships between these variables and 24-h urine composition. Patients were included from four metabolic stone clinics in three geographically distinct locations (Boston, MA and Lebanon NH; Nashville TN; and Scottsdale, AZ). A multivariate analysis was performed to examine whether a relationship exists between temperature, humidity, season of year and 24-h urine composition.

We found that increasing the temperature quintile was significantly associated with increasing urine calcium, and increasing super-saturation of calcium oxalate and calcium phosphate. Furthermore, increasing temperature was also associated with decreasing urine sodium and no changes in urine volume or urine urea nitrogen (Table 2), suggesting that urine volume, dietary sodium or dietary protein do not account for our findings. When analysing season of year, summertime 24-h urine collections were associated with increased volume relative to collections during winter. No other 24-h urine constituents appeared to be related to season of year or quintile of humidity in our analysis. We performed a sub-analysis of the sites located in the northeastern USA to confirm that the effects we report were effects of temperature itself, rather than of geographic location. The results of this sub-analysis mirrored exactly the findings from the entire cohort, suggesting that the increase in urine calcium was indeed the result of variations in temperature in our multivariate model. We also found in this sub-group analysis that urine oxalate increased as temperature increased – we are unsure of the mechanism underlying this relationship.

There are several plausible explanations for our results. One is diet – it is possible that patients consume increased calcium in hotter weather, which could account for its increased presence in the urine. Another is bone-turnover or calcium haemostasis – it is possible that the body's internal regulatory mechanisms may be affected by temperature, and that hotter temperatures may increase the excretion of calcium. Previous studies have shown that serum vitamin D and urine markers of bone-turnover are higher in summer than winter months [11]. Interestingly, in our study, a decrease in urine volume was not observed with increasing temperature and an increased urine volume was observed in summer compared with winter – although the mechanism is unclear, it is possible that increasing temperatures and sweating may lead to increased thirst, causing patients to increase their fluid intake with hotter temperatures. Although it is well known that increasing fluid volume is protective against kidney stone formation (and therefore, decreasing fluid volume increases stone risk) [4], the relationship between ambient temperature and kidney stones does not appear to be mediated through urine volume in our study. Because this is a retrospective study, we did not assess any of these potential mechanisms directly – nonetheless, our results are highly significant and suggest that the observation of increased stone formation in hotter climates or at hotter temperatures may be related to increased urine calcium and oxalate excretion.

Our study has several inherent weaknesses. It is retrospective and so subject to the shortcomings of a non-prospective study design. We used average temperature and mean humidity for our analysis – however, we are unaware of the variations between individuals in time spent outdoors and exposure to these variations in climate on the days of their urine collections. Patients were on self-selected diets and therefore, although we were able to control for various aspects of dietary intake on multivariate analysis (i.e. urine sodium, phosphorus, urea nitrogen), this could theoretically affect our outcomes if a dietary factor not included were responsible. One could better discern whether these phenomena are secondary to these self-selected diets or are truly metabolic changes in response to the warmer climate by repeating this evaluation with the patients on controlled diets. However, the retrospective nature of our study did not allow this and further investigations may therefore be warranted. Although we did not analyse stone composition in this study, we assume that the distribution of stone composition for our patients is similar to that in other reports of stone epidemiology in the USA – namely that 80–85% of stones are calcium-based. Finally, patients analysed were stone formers only and so the results may not be applicable to persons without a history of nephrolithiasis.

Temperature is significantly associated with changes in urine composition – specifically increases in temperature increase urine calcium, and super-saturations of calcium oxalate and calcium phosphate. These relationships were independent of season of year and humidity. Further investigations are necessary to understand the mechanisms underlying the relationship between increasing temperature and urine calcium excretion.

CONFLICT OF INTEREST

Brian H. Eisner is a Consultant for Boston Scientific, ACMI and PercSys. He is the owner of the Ravine Group. Mitchell R. Humphreys is a Consultant for Boston Scientific and Lumenis.

Ancillary